Monday, August 15, 2011

Book Review (#28 of 2011) Slapped by the Invisible Hand

Slapped by the Invisible Hand: The Panic of 2007 (Financial Management Association Survey and Synthesis) by Gary Gorton.

Gorton is a Yale Professor of Management and Finance and an expert in the history of banking and financial markets. Perhaps no one has done as much work on the makeup of the "shadow banking" system as he has.  This book is a compilation of a few of his papers, one of which was famously presented at the Fed's Jackson Hole conference in 2008. Gorton's work is at the top of Ben Bernanke's financial crisis reading list.

What is a bank--can you define it?  If banks are regulated, but certain non-banks engage in similar activities as a bank, what types of issues arise?  I step away from all the books I've read on the issue and come away with two definitions of a bank:
1. An entity that creates a riskless, information insensitive, liquid asset for a customer and funding that asset by creating a relatively risky, information sensitive, illiquid liability for itself. 
2. An entity that borrows short and lends long. 

Gordon gives some insight into earlier U.S. banking panics that I was unaware of. In 1907, bank consortiums facing a run basically united behind the castle walls of a clearinghouse. Deposits at any individual bank could not be redeemed and customers were instead given a makeshift loan from the clearinghouse.  In a panic no one knows which bank is safe, therefore even the healthiest ones are at risk of failure, so the clearinghouse protected all the banks until the panic subsided. Creation of the Federal Reserve eliminated this local clearinghouse concept as the Fed became the clearinghouse, and later allowed banks to fail en masse. 

Today, the repo market acts as the banking system for large firms. This activity is performed largely by non-banks and allows firms to earn interest on their large amounts of cash in exchange for collateral-- usually Treasury or other AAA assets.  This collateral is generally information insensitive, liquid, and riskless. 

"Repo trading has been likened—by repo traders—to speed chess; that is, there is no time to do due diligence on the collateral offered and indeed almost none is done."

As the repo market grows, so does demand for AAA collateral. As subprime lending increased, the collateral more and more became claims on CDOs backed by mortgages.  Once doubts began to arise about the worth of that collateral, the repo market began to seize up.  There became a general panic about which assets were safe, everyone started to demand larger haircuts and Treasuries.  As lending seized up, firms like Lehman Brothers went under and the dominoes fell. 

If you have to read one book explaining the Panic of 2007, this is the one. Gorton isn't providing much narrative here, he is an academic looking at the data. As such, much of it is fairly technical. He delves into the prospectuses of various ABS and details how CDOs operate. He looks at econometric analyses to try and develop a theory about why banks and non-bank entities make the decisions they do under regulatory regimes.

If you piece together a narrative, it looks like this:
1. Deregulation of the financial sector led to increased competition and lower profits for banks who were now competing with non-banks. 
2. "(I)n attempts to maintain profitability, banks enter new activities which are not necessarily a source of public policy concern per se, but become entwined with traditional banking activities and, hence, a source of concern." Capital requirements were navigated by doing off-balance-sheet activities and increasing securitization. 
3. As this continued, the activities of banks and non-banks were increasingly "off the radar" of bank regulators. The repo market was one example of this, the Fed stopped trying to measure the activity altogether.
4. Eventually, ABS and CDOs were created from loans that required housing prices to keep going up in order to be profitable. A sub-prime loan was made with an initial 2-3 year fixed rate that then adjusted upward, with the expectation that in 2-3 years the borrower would refinance the loan having earned equity as the value of his house increased; this could be repeated indefinitely.  The loan was sold and securitized as part of a bundle, which was then sliced and diced again into CDOs, CDO-squareds, etc.
5. In 2006, the ABX was introduced, for the first time introducing a price index for subprime securities. People could then bet/trade on what they thought they were worth, and the ABX started to decline as the market started to short subprime.
5. When housing prices stopped going up and the ABX started going down, the CDOs started to take major losses.
6. This triggered even more losses on CDS issued on the CDOs. 
7. Adherence to mark-to-market accounting rules forced major write-downs of firm assets, instantly causing many firms to become insolvent. 
7. Panic ensues. 

Gorton empirically examines two common hypotheses of the crisis and finds them wanting:
1. Lender's increasingly lower lending standards created the problem.  This isn't backed up by the empirical evidence.
2. Originate-and-distribute (securitization) created misaligned incentives that created the problem. Gorton points out that all the originators suffered major losses, many going bankrupt. All originators kept a great deal of risk on their books. 

The problem was, basically, the creation of assets that were a one-way bet on housing prices.  Once assets created from those started to rated well by rating agencies and considered riskless collateral, the system quickly spread soon-to-be-toxic risk all over the world. (I take issue with people who in hindsight say "everyone knew housing prices would fall," because if everyone knew that they wouldn't have made the loans and willfully taken billions in losses and gone bankrupt, would they?)

Gorton takes issue with a few recommendations in regards to regulation.  He sees higher capital requirements as only effective in shrinking the banking sector, which may not be what a society really wants or needs.  He also notes that until "banking" is better defined, we're always going to have a hard time policing banking activity.  Better to create a system that aligns incentives and eliminates information assymetry problems.  Mark-to-market accounting adherence during a crisis is ill-advised as well. 

There aren't any grand political statements in this book, or vilification of demons. Mostly facts and analysis and some stylized examples. 

I only give it 4 stars out of 5 because a lot of the information is jumpy and overlapping. Some of it, like his letter at the end to someone in 2107 was also a little...odd.  I also read it on the Kindle and some of the equations and charts were hard to read in Kindle format. But I highly recommend the book.  

No comments: